Aleut Grammar

Morphology — Aleut Grammar — 47
2. Morphology

2.1. Inflection and word classes

2.1.0. Survey

The open word classes are nouns (2. L 1 .) and verbs (2. 1 .9.). Ordinary nouns
have suffixes for number (singular, dual, plural), relational case (absolutive, rela-
tive), and grammatical person, so-called possessive suffixes (anaphoric third per-
son, first, second and reflexive third person). Verbs have mood and tense suffixes
(including a zero), several of which share with nouns suffixes for number, relational
case and grammatical person. Many stems are ambivalent, both nominal and verbal
(see 2.2.1.1.). There are no adjectives other than verbal nouns and participles.

Other word classes are personal pronouns (2. 1 .2.); contrastive terms (2. 1 .3.);
quantifiers (2. 1 .4.); numerals (2. 1 .5.); positional nouns (2. 1 .6.), which have the func-
tion of postpositions; demonstratives (2.L7.) and interrogatives (2,1.8.). The latter
three classes include verbal and other derivatives. In addition, there are a few par-
ticles of various types (2.1.10.).

2.1.1. Ordinary nouns

2.1.1,0, Introduction

The suffixes (2.1.1.1.) have several variants, distributed locally and over
time. In the course of time, especially in the speech of persons bom around 1930 and
later, the system has suffered certain important reductions. The stems (2.1.1 .2.) end
in a short vowel or in a consonant. In later times, the consonant stems have more and
more become stems in -i-, in a way to be explained below. Some nouns, among them
proper names, have certain peculiarities (2. 1 . 1 .4,).

2.LLLThe suffixes

In the tables below, a consonant in parentheses, (it), (g) or (ng), is used after
a vowel stem, while a vowel in parentheses, (i), is an auxiliary vowel after consonant
stems; cf. 1.3.4.1. and 4.

2. 1 . 1 . 1 . 1 . Simple number and relational case
sg, du. pL

abs. -(«) -(i)x E, Au -(i)n, A -(l)s

rel. -(i)m -"-

The plural -s at Veniaminov's time (1846:xiii) was used also at Unonak, e.g.
lagis 'Canada geese' (sg. lax); asxus 'nails' (sg. asxuj^). Later it was replaced by the
eastern -n but is still preserved in some place names, e.g. Ibxtis (1976-1982) beside
Ibxtin (1984) (Umnak 171 in Aleut Dictionary, Appendix 9; other examples are
Umnak 18, 70, 172, 185, 195). In personal names recorded by the Billings Expedi-
tion 1790-1792 (in Bergsland 1997) the -s is attested for Islands of Four Mountains
as well as for the Andreanof Islands, from Amlia through Ilak Island. The usage of
the ancient Rat Islanders is unknown. The Attuan -n might possibly have been intro-
duced from the East (03,; Aleut Dictionary, pp. xxiv f.).

48 — Aleut Grammar — Morphology

2. 1 . L 1 .2, Anaphoric third person (3 A)

sg^

du.

pi.

abs, -a

-kix

E, Au -(ng)in, A -(ng)is

rel. -(g)an

-kin

-(ng)in

These suffixes refer to a preceding term, specified in the relative case or
known from the context or the situation, e.g. taya^-m ula-a 'the man's house',
ula-a 'his house'. They may indicate the number of the possessum or (and/or) that
of the possessor (the referent), depending partly upon whether or not the latter is
specified in the construction; see 3.1.1.6,

In Eastern Aleut the relative sg. suffix is - and was - used conunonly also for
the dual and plural, the number being indicated in the following temx The dual
suffix -kin, well known from Atkan, is barely attested in Eastern, e.g. Unmak 1910
amakux lakin adakin 'toward those two sons of his' (J 51:21 original cylinder
record, confirmed at Nikolski 1982 and 1987; in Jochelson's edition 1923 altered to
lakix). In his translations of about 1870 Shayashnikov a few times used the absolutive
suffix -kix instead of the relative one, e.g. ulugakix Kugaan 'on His cheeks' (John
19.3; A 1860 ulugakin kukin), a usage found also by Marsh at Nikolski about
1950. In a few cases Shayashnikov (and Marsh's consultant) used likewise the plu-
ral suffix, which in Eastern Aleut has no case distinction, e.g. inim inkamaagungin
kungin 'on the sky's clouds' (Mark 14.62; A 1860 kuyuudam inkamiigungin
kungin). In the singular, the case distinction is generally observed but already in
1935 the absolutive could be used instead of the relative: ulam chimaluu (for
chimalugan) angadaan 'from the top of the floor of the house'.

In the speech of young Atkans, the set has tended to be reduced to a singular
and a plural without case distinction: sg. -V (vowel length, the postvocalic variant of
the abs,sg.), pi. -kix. With the young speakers (as also at Nikolski), the dual has
become obsolete as a category, while the 3 A pi. is used as a Lp.pl. (see 2.1.1.1.3.
below).

2,1.1.1.3. Possessor first, second and reflexive third person (3R). The num-
ber of the possessum is indicated only with a singular possessor.
Possessor sg..

possessum

sg-

du.

pi.

l.p.

-(i)ng

E king

A, Au ^k

-ning

2.p.abs.

-iin

-kin

E -txin, A, Au

rel.

E, Au -

min

yy )»

A -mis

-kimis

>)

3R abs.

-iin

-kin

?5

rel.

-iim

A-

kiim "-

?9

Morphology — Aleut Grammar — 49

Possessor dual plural

l.p. = plural -(i)ngin,A -mas

2.p. abs. -dix, Au -yix E -chi(n ), A -chix, Au

-chi(x)

rel. E, A -(i)mdix E -(i)mchi(n), A -(i)mchix

3R abs. -dix, Au -yix E -din, Au -yin, A = dual

rel. -max E, Au -mang, A = dual

E also -(i)mdix -(i)mdin

For the l.p. and for the sg. 2. and 3R pi. possessum there is no case distinc-
tion in any dialect.

In Eastern Aleut, the relative sg. possessum with a sg.2. or 3R possessor is
used regularly also for a dual or plural possessum, the number being indicated by
the following term, e.g. 1870 chamin ilkin 'in your hands' (Luke 43.46; A 1860
chakimis ilkin ). In Eastern the relative dual suffixes are not attested at all, and the
pi. -txin in relative function is rare in the material: 1870 adatxin Aguuguu 'your
fathers' God' (Acts 7.32); Unalaska 1910 ilaasatxin hadangin 'to his relatives' (J
34:86 cylinder recording). [The relative function of -txin, common in Atkan, was
probably an Aleut innovation. In Eskimo the plural forms in question are largely
merged with the respective singular forms.]

In old Atkan, the sg.2.p. abs. sg. had the archaic postconsonantal variant -in :
1838 sagimagin 'your face' (Matthew 6. 17); likewise 1860-1952 ugin beside ugiin
'your own' (see 2.1.2.2.). In late Atkan, the suffix -Vn (the postvocalic variant) is
only 2. p., while the 3R rel. -Vm is used for both cases. Thus, with the dual lost, late
Atkan has a case distinction only in the 2.p. sg. possessum.

The 1 .p. pi. suffix -(i)ngin, the same as the Eastern and Attuan 3 A pi. suffix,
is known from Eastern Aleut, regular in Attuan and common in modem Atkan, with
the variant -ngis (the Atkan 3 A pi. abs.). In the Eastern Aleut texts, from the 1830's
on, one finds constructions of the typetumaniin ada& "for us father", in the modem
language simplified to tuman ada^ (cf. 2.1.2.1.4.); see also 2.1.2.1. and 2.1.6. The
pronoun tuman 'we, us' contains the ancient suffix -man, which corresponds with
the Atkan -mas, as in adamas 'our father', still used by Atkan speakers, with a
variant -mis found also at Bering Island. This suffix is found also in a legend re-
corded at Nikolski in 1909: Uyumas ii! 'Our brother!' (said by a woman, J 38:15
and 42: 13), in the Attuan translation Huyungin!

The 2.p.pl. suffix E -chi is attested since 1792, the later more common vari-
ant -chin since 1860; the addition apparently is the plural -n. The Atkan -chix, with
an addition possibly from the dual, is attested since 1780. Attuan in 1909 had -chi,
in 1952 -chix, probably from Atkan.

The 3R rel. suffixes -max and -mang in Eastem Aleut were replaced, in the
first place, by -(i)mdix and -(i)mdin, apparently derived from the abs. forms, -imdix,
attested since 1839, has recently passed also into Atkan. In Atkan, the old dual forms
are used also for the plural, attested since 1860. In Eastem Aleut, as attested since

50 — Aleut Grammar — Morphology

1860, the 3R pi. forms have been replaced gradually by the 2.p.pl. forms, e.g., with
the suffix in both functions, 1870 chachi kumchi aisix 'they (shall) lay their hands
on you' (Luke 21.12). By the 1950's, this usage had been adopted also, partly, by
Atkan speakers (e.g. in N.M., beside the old usage).

2,1.1.2. Types of stems

2.1.1.2.1. Stem vowel vs. consonant

A nominal stem may end in a short vowel or in a consonant. The difference may
show up in the simple absolutive singular, where a vowel stem has the suffix -5, a conso-
nant stem possibly a zero, e.g. ula-* 'house' (pi. E ulan, A ulas) vs. E ula^ 'bear-
berry' (pl.ulagin).

Another diagnostic form is the 3A abs.sg., where the suffix -a is simply added
to a consonant stem but assimilated to the vowel of a vowel stem, yielding a long
vowel (1.3.1.), e.g. haniji 'lake', 3A sg. hanii, vs. alii 'old man', 3A sg. aliga; ului
'meat; corpse', 3 A sg. uluu, vs. ului 'hatch, cockpit (of baidarka)', 3 A sg. uluga.

Before suffixes with an auxiUary vowel i or an initial long ii, a consonant stem
does not differ from a vowel stem in i, cf. anax 'club', A pi. anagis, sg.3R sg.
anagiin 'his own club', and angalii 'day, daylight', pL angalis, sg. 3R sg. angaliin.
Through such forms the consonant stems have more and more, in the COMTSe of time,
become vowel stems, leaving only a small residue, if any, in the speech of the younger
generations (people bom, say, after 1920), e.g. anax 'club', 3 A sg. anaga, andanagii
(E 1834-, A 1952-), anagii (E 1909-); isui 'harbor seal', pi. E isugin, A isugis, sg.
later also isugii^ (E, A 1978-); kanuui 'heart', 3A sg. kanuuga, and kanuugix (E
1834-, A 1949-, Au 1949-1952), kanuugiL The elimination of the consonant stems,
which meant a considerable simplification of the nominal inflection, apparently went
on differently in the different grammatical forms and differently for the different
types of consonant stems. The difference between consonant and vowel stems thus
has been a gradual one over a long period of time.

2. 1 . 1 .2.2. Stems with a final velar or uvular fricative

The great majority of consonant stems end in a velar or uvular fricative,
before the 3A sg. -a, rel. -an, mostly also in the simple abs.sg., see above. Addi-
tional examples are E lax 'Canada goose', pi. lagin (sg. also E 1909-, A I860-, Au
lagi£); qigux 'loon', rel. qigugim (abs.sg. also E 1840-, A 1948-, Au qigugij^); uliix
'boot', du. uliigix, 3A sg. uliiga (E 1834- also abs.sg. uliigi* ); E 1878-, A 1840-
1878 hiix 'oystercatcher', pi. E hiigin, A hiigis (sg. also E 1840-, A 1952-, Au
hiigii^); E nui 'stone, rock', 3 A sg. nuga, pi. nugin (abs.sg. also 1834- nugi^^).

In the 3A plural, the consonantal inflection coincides, in the absolutive case,
with the simple plural and appears to have become obsolete earlier than in the 3A
singular. The few Eastem Aleut examples are found in the biblical translations, e.g.
1838 Israayilam tanadguchiigamagin, 1870 L tanadgusiigamagin 'the cities of
Israel' (Matthew 10.23); 1870 husum kanuugin 'the hearts of everybody' (Acts
1 .24; A 1 838 kanuugingin with final n for s). About 1950 Marsh recorded in Nikolski

56 — Aleut Grammar — Morphology

'in the field, out in the field, to the camps' and an ablative, E 1838-70 tanasxaan, A
1860 tanasxaax 'from the field; out of the country'. Cf, 2.1.6, Positional nouns.

2. L 1 .4.3. Proper names

Names ending in a nasal like Suung (headland, Atka 566), Kiigun (Cape,
Atka 640, obsolete abs.sg., 1840- kiiguusii^ 'mountain'), En 1909 Ugdusxiin (a
man, J 44:8 ff.), have no special relative form but are used also in a relative function,
e.g. Suung ixchxa 'the neck (isthmus) of S.' (Atka 568). Likewise Russian personal
names such as Imaan, Ivaan 'John', Filiip 'Philip'. The personal name En 1910
Kanga^simaaj^ was used in this form both in an absolutive and a relative function,
in the latter function beside Kangai^simaagim (J 74:1, 31 and 21, etc.).

Names borrowed from Russian or English with a final long vowel are treated
like 3A sg. forms, e.g. Andrii, rel. Andrigan 'Andrew' (Russian Audrey); A 1979
Kilisnugan hadan 'to Killisnoo'.

Personal names, however, also admit functional possessive suffixes, e.g. A
1979 Taatakuchaj^ hamay / Rumaaniin / hingay uhlii agiitana^ '"Little Dad"
(nickname) lived together with his Roman only'; Hingaan aslaan hawan (for
hawakus) Viirangis ama Aalisingis hamang anagulas 'At that time those (daugh-
ters) Vera and Alice of his were not there.'

In Atkan, names of islands (countries) with a plural possessive suffix indi-
cate people of the place, e.g. Umnax (for Unmax), 1973 Umnagingis, rel.
Umnagingin 'the people of Umnak'; 1981 Amlax, Amlagingis 'people of Amlia'
(cf. Amiagim angaginangis 'the people of Amlia'), Amlagidix agiitaasal ayuxtanas
'leaving their people from Amlia they set out'. Salamatov 1860 wrote, for example,
Galileeyangis akuxt 'you are a Galilean' (Mark 14,70), but alsoludeeyangiziin 'to
the Jews' (John 8,31) rather than "^ludeeyangimin (-ngiz- abs., -ngin- rel. + (ng)iin ).
The Eastern forms in -(ng)iin look like ablative forms (cf. 2.1.6.), e.g. 1834
Akutanangiin Akungangiin 'the people of Akutan (Akutana^), the people of Akun
(Akungan)' (V A 10:6), 1909 Alaxsxiin 'the people of the mainland (Alaxsx-a )'.

2.1.2. Pronouns

2. 1 .2. 1 . tx(i)- / ti- personal pronouns

This stem, a "demonstrative of speech", represents the person referred to by
the suffix, namely a first, second or reflexive third person, in the absolutive case. As
free forms these pronouns are used primarily as an object, like a fully specified
noun. As enclitics, they are subject markers (in the moods and tenses specified in
2.1.9.), for persons other than a non-reflexive third person (which in the respective
moods and tenses are marked by simple number suffixes); for the non-reflexive third
person there are no personal pronouns other than demonstratives (2.1.7.). In addi-
tion, the free forms, in Eastern Aleut with or without an element -(ng)anaan, may
double a following person marker, as a more or less emphatic subject or, in Eastern
Aleut, possessor. The 1 .p.pl. pronoun has its peculiarities, cf. 2. 1 . 1 . 1 .4. In Atkan, the
stem admits also certain derivative suffixes.

Morphology — Aleut Grammar — 57

2.1.2.1.1. Object forms
l.p. sg. ting

du.-pl. E 1791 tximan, 1804- tuman, Eb 1984 tumin; A1780 tximas,
1838- timas, 1950- also tingin, tingis, 1987 timis; Au tingin
2.p. sg. txin, A 1950- also tin; Au tin

du. txidix, A 1973- also tidix; Au tiyix
pi. E txichi, 1878- also txichin; A txichix; Au tichi
3R sg. txin, A 1950 also tin, 1973- also tiim; Au tin
du. txidix, A 1973- also tidix; Au tiyix

pi. E txidin, 1870- also txichi(n) (= 2.p.pL); A txidix (= 3R du.),
1950- also txichix (= 2.p.pl.); Au tiyin
As the object of a verbal form with the same person as subject, all the forms
are reflexive, e.g. A ting achixakuq *I am teaching me (myself), I am learning', cf.
ting achixaku^t 'you are teaching me'. In the dual or plural the relation may be
reciprocal, e.g. A sunax txidix hnukux 'the two ships reach each other, meet' ; txidix
yaxtakus 'they love each other'. In Eastern the l.p.pl. tuman is reflexive with a
verb in the passive (cf. 3.1.1.2.), e.g. Eu 1909 ... ngaan tuman ayugnilgalix ... 'we
started on him' (J 34:27), cf. Ea 1910 ngaan txin ayugnilix 'he moved to it' (J
28:17).

The number of the object and the subject may be different, e.g. A 1952
txidix hikugaan 'when he called his people', lit. 'them(selves)' (N.M. 3:91).

A personal pronoun may also serve as a predicate noun, e.g. A 197 1 ting SkU
'it's me (in the picture)'.

2.1.2.1.2. Enclitic subject

The personal pronouns other than the l.p.pl. are added to the 3.p.sg. in the
tenses where a 3.p. subject is marked by a simple number suffix, e.g. A (asxinu^)
hilakuj^ 'she (the girl) is reading', hilaku^txichix 'you (pi.) are reading'.

Together with a final uvular the pronoun ting becomes -qing, in Atkan and
Attuan apocopated to -q, e.g. A hilakuq(ing) 'I am reading'. Together with the
enclitic negation +ulax it became in old Atkan +ulak (modem analogical +ulaq). In
Atkan and Attuan the 2. and 3R sg. txin, tin in enclitic position is apocopated to -t,
e.g. A hilakui^t 'you are reading'.

The 1 .p.pl. pronoun is used as an enclitic subject only in Atkan, in the con-
junctive, e.g. 1860 alqui maakatimas 'what can we do ?' (Luke 3.10).

2.1.2.1.3. Free forms as subject

A free pronoun apparently is necessary in subject phrases like the follow-
ing: A 1860 Ting ama Adai^ ataqan akus, E 1870 Ting kayux Adang ataqan
akun 'I and the/my Father are one' (John 10.30); A 1978 ting asxuunulax txin
satxa& tagaagan a^s 'who is going to check the gill net, I or you?' Likewise in
connection with the particle kayux, modem A kay 'also, too', e.g. Eu 1909 Ting
kayux agiitaakingin maayuqaning. 'I also prepared to go with them,' (J 40: 17); A

58 — Aleut Grammar — Morphology

1971 ting kay qaatukuq 'I, too, want to eat (am hungry)' (in idiomatic Atkan
qaatumakuq).

As an emphatic subject, the 1 .p.sg. and 2.p. pronouns have in Eastern mostly
an enclitic -(ng)anaan or -ngaan (perhaps syncopated), e.g. E 1870 (and Ep 1983)
Taga Tinganaan imchi tunuj^takuqing: 'But I say to you:' (John 4.35; A 1860
without a free pronoun); but also Wayangin imchi Ting hikuning, A 1860 (and
1971) Wakus Ting imchi hii^takuning, ''These things I have spoken unto you,'
(John 15.11); E 1870 taga Txinanaan alqutan hiisaduukakuun?, A 1860 alix
Txin alqus kungin hiimis agikutxin? 'but what will you (versus the law of Moses)
say about it (A pi.)?' (John 8:5); E 1860 taga txichingaan (1870 txichinganaan)
kiin Ting achSiku^txichi?, A 1 860 alix txichi kiin Ting ataasatxichi? but whom
do you say that I am?' (Mark 8.29); En 1948 txichingaan amaanuuchi 'go away all
of you', cf. sg. amaanuda 'go away'.

For the 1 .p.pl. the older Eastem usedtumaniin, the later tuman without the
enclitic, irrespective of the form of the verb, e.g. E 1870 taga tumaniin alqutan
maa^tan? (= 3,p.pl.), A 1860 alix timas alqu^ maai^tas? 'but what shall we do?'
(Luke 3.10 resp. 13); E 1870 tumaniin Missiiya^ ukulgakui^ (passive), A 1860
Messiiya* ukukus (=3.p.pl.) 'we have found the Messias' (John 1.41); Ep 1984
tuman aguqangin (= 3.p.pl.) 'we made it'; En 1983 tuman kayux saamhlalgadakun
(passive pi.) 'we, too, usually gather eggs'. The A 1860 timaziin in the following
sentence probably was an imitation of Eastem: Taga timaziin Hingan ida^tazulax
(= 3.p.pl.), E 1870 tumaniin haqatalgakui (passive) 'But we know Him.' (John 7.27).

2.1.2.1.4. Free forms as possessor

In a phrasal referent a free pronoun is found both in Eastem Aleut and in
Atkan, e.g. E 1870 txichi kayux tumaniin quchxingin, A 1860 timas ama txichi
quchigmas 'between you and us (A us and you)' (Luke 16.26).

The doubling of a possessive suffix with a free form is an Eastem Aleut
usage (forAtkan see 2.1.3.1.), e.g. 1870 ting anging 'my spirit' (Luke 1.47; A 1860
without ting); E 1871 tinganaan imlining 'my hair'; En 1975 ting latuging 'my
grandfather'. The following Atkan phrases are probably imitations of Eastern: 1840
( 1838) timas umsumas 'our tongue', E tuman (1870 tumaniin ) agna^ (Acts 2.1 1);
1860 timazaan Adamas 'our Father', E 1870 Tbmaniin Ada^ (Luke 11.2). As
mentioned in 2.1.1.1.3., in modern Eastem tuman is used without the enclitic, e.g.
Ep 1984 tuman sunai^ 'our ship', tuman sunan 'our ships', both also tuman
sunangin ; Ek 1982 tuman tana^^ 'our island' (not tanangin).

2. 1 .2. 1 .5. Derivatives

In Atkan, certain derivative suffixes may be inserted in the free forms, e.g.
1979tiiguzangukw^taqadaa^t 'quitlooking at me specifically!'; 1973tiidahlingin
tingin igatal ... 'even we (not only the old people) became afraid' (changed from
uudahlingin, cf. 2. 1.3.3. u-hli); 1987 tii^siidang 'poor me' ; txikuchaan 'you little
one'; t(x)iikluun 'you clumsy one'; tina&chi^ikuchaan 'you little darned one'.

Morphology — Aleut Grammar — 83

rogative, or after any other term put in question, having everywhere the rising tone,
e.g. A Qilagan Piitrai hla^ tugal saganai^ ii? 'Did Peter knock the boy yesterday
(or maybe he did not)?'; Qilagan Piitrai hla^ tugal ii? saganai^. 'Did Peter really
knock the boy yesterday (or maybe just scolded him)?'; Qilagan Piitra^ hla^ ii?
tugal sagana^. 'Was it the boy (or somebody else) that Peter knocked yesterday?';
Qilagan PiitraS ii? hlaii tugal sagana^. 'Was it Peter (or somebody else) who
knocked the boy yesterday?'; Qilagan ii? Piitrai^ hlait tugal sagana:^. 'Was it yes-
terday (or some other day) that Peter knocked the boy?'.

Questions with one of the interrogatives listed above have the same intona-
tion as a declarative sentence. With a final particle with rising tone the question
becomes a question about the question, e.g. A Kiin haqal ii? 'Who came? (is that
your question?)',

2.1.9. Verbs

2.1.9.0. Introduction

Verbs differ from nouns by having mood/tense suffixes. The so-called Gen-
eral (2. 1 .9.3. 1 .), however, is unmarked, and many stems are ambivalent (see 2.2. 1.1.).

Like nominal stems (2.1.1 .2.), a verbal stem may end in a short vowel or in
a consonant: a velar or uvular fricative, a labial nasal, or -t-, which alternates
with -ch- or -s- before the initial vowel of certain mood/tense suffixes (1.3.3.). Be-
fore certain mood/tense suffixes a consonant stem has an auxiliary vowel, which in
the later language has largely become a stem vowel (1.3.4.1., cf. 2.L1.2.1.), or a
mood/tense suffix may have a postvocalic and a postconsonantal variant (separated
by I below; cf. 1.3.4.3-4.).

Most of the verbal forms may come at the end of a sentence final (main)
clause as well as at the end of a non-final (subordinate) one; some of them are used
also in phrases with certain auxihary verbs (2.1.9.10.3.; 3.8.). Two moods (2.1.9.6.
Imperative and 2.1.9.7. Prohibitive) occur only in a final clause, while two other
ones (2.1.9.8. Anterior and 2.1.9.9. Conditional) are primarily non-final.

In general, verbs with a fiiUy specified or no complement are marked for
the subject only, while a double marking includes anaphoric reference to a 3.p.
complement (see 3.2. ff.). In many cases a simple subject is marked by a number
suffix (2. 1 . 1 . 1 . 1 ) or an enclitic pronoun (2. 1 .2. 1 .2.), while the anaphoric reference is
marked by possessive suffixes (2.1.1.1 .2-3.). In other cases there are special suffixes
(including possessive ones) for the subject and/or for the anaphoric complement.

The negation is in some cases suffixal, preceding or combining with the
mood/tense suffix, in other cases enclitic, possibly followed by the enclitic subject
pronoun.

In certain cases, to be discussed in the syntax, a non-final verbal form has in
addition an enclitic dative: a 3 A form of the positional noun i- (2.1.6.1.3.), sg.
+(ng)aan, Au +aa, + anaa, du. +ikin, pi. H-(ng)iin.

84 — Aleut Grammar — Morphology

2.1.9.1. -(i)ku-, neg. -lakag- (late A -laka-) Present

E.g. as^-kuii 'he/it died' , as^-laka^ 'did not die* ; ayug-iku& 'he went out (in
his boat)' , ayug-laka^ *he did not go out' ; (taya|ui) as^t-ikui 'he killed (the man)' , E
as^alaka^ (deletion of t before 1), A asi^aMakaJ^ 'he did not kill (the man)'.

Present forms are used as (1) sentence final predicates, (2) non-final (subor-
dinate) predicates, (3) referent, in the relative case, of certain positional nouns.

2.1.9,1.1. Sentence final predicates
Forms with a subject marker only

sg.

du.

pi.

3.p.

-ku-i

-ku-x

E, Au -ku-n, A -kt

l.p.

E -ku-qing A, Au

-ku-q

= pL

?»

2.p.

E -ku-^-txin

-ku-]^4xidix

E -ku-i^-txichi(n)

A -ku^t, Au -kut

Au -kutiyix

A -kui^txichix
Au -kutichi

3.p.

-lakai^

-lakagix

A also -lakax

E -lakagin

A -lakaps, -lakas

l.p.

E -lakaqing
A, Au -lakaq

= pL

»»

2.p.

E -lakaji-txin
A -lakai^t

-lakai^-txidix

E -lakai-txichi(n)
A -iakajctxichix

Forms with anaphoric reference to a complement (3. p.)

subject 1 anaph. sg.

du.

pi.

3A

-kuu (-ku-a)

-kukix

E, Au -kungin
A -kungis

l.p.sg.

-kung

E -kuking
A-kuk

-kuning

2.p.sg.

*kuun

-kukin

E -kutxin

A, Au -kut

anaph. sg./du,/pl,
l.p. du-pl.. (E passive) A -kumas, -kungin, -kungis, Au -kungin
2.p.du. -kudix, Au -kuyix
2.p.pl. E -kuchi(n), A -kuchix, Au -kuchi(x)

anaph. sg.

du.

pi.

3A -lakaga

(not attested)

E -Iakag(i)ngin

late A -lakaa

A -Iaka|is, -laka(gi)ngis

l.p.sg. -lakaging

A -lakagik

-lakagning

late A -lakang

2,p.sg. -lakagiin

A -lakagit

old A -lakagin, late -lakaan

Morphology — Aleut Grammar — 85

anaph. sg./du./pl.
1 .p.du-pl. A -lakagmas, Au -lakagingin
2.p.pl. A -laka^^chix

In general, a statement in the grammatical present refers to a state-of-affairs
that obtains (or does not obtain) at the time of speech, or to a process or an event that
takes place (or does not take place) at the time of speech, or took place (or did not
take place) a short time before the act of speech, e.g. ulai^ quhmakui 'the house is
white'; ulai tagadalaka^ 'the house is old (lit. not new)'; adang sagakui^ 'my
father is sleeping'; A Viira^ hilalakai 'Vera is not reading (but doing something
else)'; Piitra* waagaku^ 'Peter is coming back' or 'Peter just came back'; Piitrai
asxu^ suku^ 'Peter is taking a/the nail' or 'Peter took the nail (right now)' ; Piitram
sukuu 'Peter took it (right now)'; Eu 1910aaliisimin ilan alaji imin kasukuqing 'I
have found a whale for you in your harbor' (J 35:52); Eu 1909 maarsalam waagaqaa
nidilim ilan txin aagatalakagim miigum boochukangin adaam aglaku^ 'the
marshal who came her has for a week without interruption been hauling barrels of
beer to his place' (J 38:8).

The present is compatible with derivational suffixes such as E -saagu-,
A -zaagu- 'recently, a while ago'. With certain other derivational suffixes and in
certain phrases it refers to a near future (3,8. 1.3.1. and 3.). It is compatible also with
the counter-factual particle kum (2. 1 . 10.4.), e.g. A Piitra^ kum ting kiduku^ 'Pe-
ter would have helped me (now, but he is not here)'.

In Eastern, and in Atkan influenced by Eastern, the present can be used also
in a question, e.g. Ea 19 10 "Ingaya umning a^taku^txin hi?" 'Are you my nephew
then («Jita- inferential: judging from what you say) ?' (J 15:49); A (E) ukukuun?
'did you find it (right now)?' In Atkan, and frequently also in Eastern, a question
referring to the present time or to the near past is put in the conjunctive (2. 1 .9.2.) or
in the general (2.1.9.3.1.)

2.1.9.1.2. Non-final predicate

The forms listed above are used also in various non-final constructions to be
discussed in the syntax, including 2.p. forms used as 3R (cf. 2. 1 . 1 . 1 .3. and 2. 1 .2, L 1-
2.; E 3R pi. *-kui^-txidin not attested). The time reference is then shifted from the
time of speech to that of the following clause.

There are also simple relative forms, sg. -kum, -lakagim, du. and pi. like
abs., used in coreference with a following 3. p. subject (3.11.1.1., etc.), e.g. A
anqa^takum haqaagan agiku^ 'he went away but will come back'.

Forms with the enclitic +ngaan etc. are 3.p.sg. -kugaan (Au *kugaa,
-kug^aa), "lakagaan, du. -kugikin, pL E -kuniin, A -kuziin (late -kuzaan), -lakag^ziin;
l.p.sg. E -kuqingaan, A 1860 also -kuqaan, later -kuqaang, Au -kuqaa;
2.p, and 3R sg, E -kuitxinaan, A -kuitaan, -laka^^taan, Au -kutaan, du. -ku&txidigaan,
Au -kutiyigaan, pi. A -kui^ichigaan; 3 A sg. -kuungaan, sg. l.p.sg. -kungaan, etc.

126 — Aleut Grammar — Syntax

3. Syntax

3.0. Introduction

Most Aleut sentences and clauses have a verbal predicate, which normally
comes at the end (qualification in 3.1. LI.)- Nominal sentences are constituted by
demonstratives (3.1.3.). In Eastern Aleut there are also nominal sentences of the
Russian type, no doubt due to the Russian impact, while the corresponding sen-
tences in Atkan and Attuan have a verbal copula (3.1.2.). The verbal predicate of a
simple sentence or of the final clause of a complex one carries the temporal and
modal marking in relation to the act of speech, while the verb of a non-final clause is
marked in relation to the following clause. A complex sentence may contain an
unlimited number of clauses, in actual texts recorded on a cylinder phonograph
(Jochelson 1909) or on magnetic tape up to some fifty. Quite often the speaker ap-
pears to have the choice between making a "full stop" (the final absolute tense and
mood marker) or adding more clauses to his sentence. An Aleut complex sentence
thus appears to be a chaining structure constituted by a "running superordination".

A simple sentence or clause may include a subject or no subject (3.1.1.3.
meteorological expressions, 3.4.3. clauses in the passive). The predicate may be a
verb with no complement, a predicate noun with a copula (3.1.2.), or a verb with a
preceeding direct object in the absolutive case (3.2.) and/or an oblique term or local
complement (3,3. a positional noun in a local case with a nominal adjunct and/or a
possessive suffix, or a demonstrative or interrogative form). The number of terms
(arguments) may be increased or decreased by verbal derivative suffixes (2.2.5.,
3.4.). In addition there may be a temporal adverbial (3.7.).

The terms (arguments) of a clause may be specified (explicit) or anaphoric.

The verb of a simple sentence or a final clause with fully specified or no
complements may have a nominal subject in the absolutive case or a 1. or 2.p. sub-
ject marker, enclitic or suffixal (3.1.1 .1-2.). If the nominal subject is left out as known
from the context or the situation the verb implies an anaphoric reference to it, e.g.
(hyphens added for clarity)

tayagu-i awa-ku-&

man-sg.abs. work-pres.-sg. 'the man is working'

awa-ku-ic 'he is working'.
There are no non-reflexive 3. p. pronouns other than demonstratives.

If a 3.p. complement or a subordinate part of it is left out as known from the
context or the situation there is in general a suffixal reference to it in the final verb
and a nominal subject is in the relative case, e.g. (Atkan)

Piitra-ii tayagu-x kidu-ku-^

Peter-sg.abs. man-sg.abs. help-pres.-sg. 'Peter is helping the man'

Piitra-m kidu-ku-u (< -a)

Peter-sg.rel. help-pres.-sg.3A 'Peter is helping him'

Piitra-* tayagu-m had-a-n huya-ku-$

Peter-sg.abs. man-sg.rel, direction-3A-loc. go-pres.-sg.

'Peter is going toward the man'

Syntax — Aleut Grammar — 127

Piitra-m had-a*-n huya-ku-u (< -a)

Peter-sg.rel. direction-3A-loc. go-pres.-sg.3A

Teter is going toward him\
If also the nominal subject is left out the verb includes a double anaphoric reference:
kidu-ku-u 'he is helping him', had-a-n huya-ku-u 'he is going toward him' . Under
certain conditions to be discussed in the relevant sections below, the referent may be
included in the sentence as an outer subject preceding the referential phrase.

A nominal term may be simple or complex, possibly including a nominal
adjunct in the relative case (in the simple dual and plural like the absolutive case)
and/or a possessive suffix. If the nominal adjunct is left out as known from the
context or the situation the possessive suffix of the head includes a reference to it
and the anaphoric reference is marked also in the final verb (qualifications in
3.1.1.6.2.), e.g. Atkan

hla-s ada-a awa-ku-$

boy-pl. father-sg.3 A work-pres.-sg. 'the boys' father is working'

ada-ngis awa-ku-s

father-3A pi. work-pres.-pl. 'their father is working'
When the adjunct is specified (hla-s), the verb agrees in number with the head of the
subject (ada-), here a singular (-x). When it is anaphoric, its number is included in
the possessive suffix of the subject (-ngis 'their') and the verb agrees in number with
the referent, here a plural (-s). Also in such cases the referent may be included as an
outer subject of the sentence (3.1.1.6.3.).

The choice between a fully specified term and an anaphoric reference to it
runs through the whole sentence structure, determining also the transition from one
clause to the following one in a syntactic chain.

3.1* Subject and predicate

3.1.1. Verbal predicate.

3.1.1.1. Nominal (3.p.) subject and verbal predicate

In general, a nominal subject precedes the verbal predicate and the verbal
predicate agrees in number with the nominal subject.

In a final predicate in the present (2.9.1.), in participial tenses (2.9.3.), and in
the optative (2.9.4. 1.), the number of the nominal subject is marked in the verb by a
simple number suffix, e.g. present A tayagu-ii haqa-ku-& 'the man is coming',
tayagu-s haqa-ku-s 'the men are coming'; remote past E ali^ awa-na-g-ulux 'the
old man did not work', aiig-in awa-na-n-ulux 'the old men did not work'; optative
A 1860 Angali-in haqa-a^ta 'Thy kingdom (lit. daylight) come' (Luke 1 1.2), mod-
em A haqa-ai^ta-^ 'may he come, let him come'.

The conjunctive (2.9.2.) has no number suffixes, e.g. A Mariiya^ hila-1 ii?
'is Mary reading ?', asxinu-s hila-1 ii? 'are the girls reading ?'. In the dubitative
(2.9.5.) and in the non-final moods intentional (2.9.4.2.), anterior (2.9.8.) and condi-
tional (2.9.9.), the number is marked by possessive 3A suffixes, e.g. dubitative A

128 — Aleut Grammar — Syntax

kiin ayi^aasii akum haqal hingama-ai^ta-a '(I wonder) whose boat (it is that) is
coining there ?';..* itiiaygi-s a*ai^ta-ngis (ukujitanaai^s) ^trying to see) if there are
reindeer (there)'.

A collective singular may have the verbal predicate in the plural, e,g. En
1909 .•* iqyamixti-x .•. iluugaqali$ta-ku-n ... *(saw) a fleet of baidarkas beginning
to come inside' (J 45:3). Similarly with a numeral or quantifier construction for-
mally in the singular, e.g. Eu 1909 Atim signa|ii tayagu-i^ Ikalgan kugaan walita-
ku-n. 'More than ten men came here from Chemofski' (J 38:34); Ukalgan angagigan
usu-u kaangulix kayux ukudiga-ku-n. 'AH the people here are well and in good
health.' (J 39:2.).

In Atkan and Attuan a plural or dual subject may occasionally have the predi-
cate in the singular, e.g. A 1909 waan tanadgusim angagina-ngis hagang al ai^ta-
ku-5 'the people of this village were apparently up there' (J 78:102); waku-x
alignamqu-x agisadix hangalix, angali-ku-x kanga[an] a§a-ku-g-aan (3.11.1.1.),
... 'the darned old men ascended their observation hill and when they reached the top
ofit,...' (J 79:14-15).

Sentences with no subject word are mostly anaphoric. For example, a story
begins as follows (En 1910): Ayaga-j^ malga-qa-j^ aj^ta-ku-j^. Ugiii-ij^ta-[j^]
ayagagusxadaagali-ku-m, ugigiitu-na-g-ulux. 'There was a woman. She was con-
stantly courted for marriage but did not want to marry.' (J 62: 1-2).

In Eastern a definite subject frequently comes last, as a repetition of a previ-
ously mentioned one (possibly a case of Russian influence), e.g. En 1910 Amaligan
alix qingaadalix, qidaqali-na-£ aman ayaga-j^. (lit,) 'Being there feeling cold, she
began to cry, that woman.' (J 55: 18; A 1973 Haman ayagaic hamahligan al achul
qidaqalina^). In Atkan this usage seems to be rare, having more the character of an
afterthought, e.g. 1952 Qichitim angunaa agunai^ a^ta-ku-^ hama-n chngatu-^,
kadim hadaan aasa-na-x a^ta-ku-i. 'It apparently made much money that sea
otter, formerly it apparently did.' (N.M. 2:50); Aagayuugis kamlixtal hamama[na]i^
a&taku^, Taakdaada^ akum. 'He apparently was out shooting cormorants, "Blad-
der Pouch" (nickname) was.' (a-ku-m subordinate sg.rel., coreferential with the main
predicate).

3.1.1.2. Verbal predicate with a 1. or 2.p, subject

In the moods and tenses where a 3. p. subject is marked by a simple number
suffix, the speaker and the person(s) spoken to are marked as subjects by enclitic
personal pronouns, the "demonstrative of speech" ti-, tx(i)- with possessive suffixes
for the respective grammatical persons, added to the 3.p.sg. form (2.2. 1 .2.), virtually
suffixes (see 3.1.1.4.), e.g. Echitaaya-ku-qing(*'ku-&+ting) 'I am reading', chitaaya-
ku-$-txin 'you (sg.) are reading'; A hila-ku-q 'I am reading', hila-ku-j^-t 'you are
reading', hila-lakaq (-lakax+ting) 'I am not reading' ; Au 1952 chumu-ku-q 'I am
stepping down in the snow', unguchi-ku-t 'you are sitting'; E awa-na-g-ul-ting 'I
did not work', awa-na-g-ul-txin 'you did not work' (cf. 2.9.3.); A qa-ai^-txichix
'please (you pi.) eat' (see 2.9.4. 1. 1.). Likewise in the prohibitive (2.9,7.), e.g. Esisa-

Syntax — Aleut Grammar — 129

a§ana-i^-txin, A sisa-agana-x-t 'don't get lost'. The enclitic person markers are
added to the conjunctive as well, e.g. A qaatu-t ii? 'do you want to eat T (deletion
of -lix); E qa-laka-txin ii? 'didn't you eat ?' (deletion of the final nasal of -lakan);
dui^taasa§i-laka-txichi hi? 'you don't have a guest, do you ?' (J 41:26).

In the dubitative, intentional, anterior, and conditional, the person markers
are possessive as in the 3.p., e.g. dubitative A txin sasuliti-ij^-ang 'do I bother you ?'
(see 2.9.5.). The imperative has a special suffix for the 2.p.sg. and possessive suf-
fixes or enclitic pronouns for the 2.p.du. and pi. (see 2.9.6.).

For the non-singular first person (the speaker with one or more other per-
sons) one uses the 3.p.pl. form or, especially in Eastern and later Atkan, a passive
form (2.2.5.3.), e.g. A awa-ku-s, E -n 'they/we are working'; optative E 1837kana^-
ai^ta-n(stemkanai^t-), A 1838-1862 kamga-a*ta-s 'let us pray'; En 1910 Asagai,
unugulux as$a-laagana-n axta-ku-n ! 'Cousin, we are already lost (lit. died long
ago) !' (J 68:39); E, A awa-lga-ku-S "there is being worked" = 'one is working, we
are working' ; optative A 1 952 ayuxta-lga-ai^ta-i^, Au ayuxta-lu-ut iet us go out (in
the boat)'. The difference between a 3.p.pl. and a l.p.pl. would show up in a following
superordinate verb, the reference to the speaker entailing no anaphoric reference.

In Eastern the l.p.pl. may be marked also by the free pronoun tuman 'we,
us', old E tumaniin 'for us', see 2. 1 .2. 1 .3. In Atkan the 1 .p.pl. pronoun is used as an
enclitic subject marker in the conjunctive (with deletion of -lix), e.g. 1860 ^Iquk
maaka-timas 'what can we do ?' (Luke 3. 10, E optative tumaniin alqutan maaj^tan);
1979 igana^ malgal ayxal gumal axta-tingis 'we were (apparently) travelling in
rough water (lit. it being terrible)'. In the optative, an indefinite construcfion (see
3.5.1.) may imply reference to two or more persons including the speaker, e.g. A
1952 ama^u-na-x a-ai^ta-x 'let us go racing (in skiffs)'.

For the use of free pronominal forms as an additional subject see 2.1.2.1.3.
For the contrastive inaqa- '-self see 2. 1 .3. 1 .

3.1.1.3. Sentences without a subject

A sentence without a subject word may have an anaphoric subject (3.1.1.1.)
but meteorological expressions have no subject, e.g.

chij^ta-ku-x

be wet-pres.-sg. (1) 'it (e.g. the ground) is wet', (2) 'it is raining'.
The difference would show up in a following, superordinate verb (cf. 3.15.1.1.2.1.),
e.g. A

(1) chi^ta-ku-x ukuxta-qa-ng 'I saw that it was wet'

(-ngsg.l.sg. 'it r, see 3.2.2.),

(2) cliixta-ku-x ukuita-na-q T saw that it was raining'

(-q = -^+ting'r, see 3.1.1.2..);

(1) haqa-a^ta-a haqata-lakag*ing 'I don't know whether he is coming'
(dubitative),

(2) ayangi-i^ta-a haqata-lakaq (4akai^+ting) 'I don't know whether it is

foggy'-

Aleut Compared With Eskimo — Aleut Grammar — 341

4. Aleut sentence structure compared with Eskimo

Aleut and the Eskimo languages — Alaskan Yupik, Siberian Yupik, and Inuit
— have in common important structural features and a large body of lexical and
granmiatical morphemes and appear to be later forms of the same proto-language
(updated views in Bergsland 1986, 1989, 1994, 1997b;Fortescue,Jacobson, Kaplan
1994). Aleut, however, has a simplified morphology and a sentence structure differ-
ing from Eskimo in fundamental respects. The following pages are a brief discus-
sion of some major points.

4 1. The case systems

The crucial Aleut innovation is the reduction of the nominal case system.

Eskimo and Aleut ordinary nouns (with or without a possessive suffix) and
most pronouns have two grammatical (relational) cases, absolutive and relative, in
three numbers (2.1.1.1., 2.1.7.3., 2.1.8.), marked by mostly cognate suffixes. Es-
kimo nouns and pronouns have in addition a number of oblique (local) cases, while
local cases in Aleut are limited to positional nouns (2.1.6.) and to demonstrative and
interrogative adverbs (2.1.7.4., 2.1.8.3.2.). Aleut has only two local cases, called
locative and ablative, but the comparison of the allomorphs shows that the locative
also reflects an allative and the ablative also a prosecutive, corresponding with the
respective Eskimo suffixes (see 1986: 1 14 f.). Certain special forms and relics show
that the oblique (local) cases of nouns were lost in Aleut, rather than being an Es-
kimo innovation.

Yupik nouns and pronouns have five oblique cases: allative (also called
terminalis), locative (localis), ablative (ablative-modalis, insimmcntal), prosecutive
(vialis, perlative, translocative), and equative (aequaUs). The Inuit cognate of the
Yupik ablative is an instrumental (modalis), Inuit having an extra ablative
(distantialis), probably an innovation (see 1989:32 f.). The oblique case suffixes are
added to the relative forms of nouns and pronouns with various assimilations. The
initial n of a case suffix is assimilated to the simple rel.sg. -m (eastern Inuit -p) into
m: allative -nun, -mun (eastern Inuit -nut, -mut); locative -ni, -mi; ablative Alas-
kan Yupik -nek, -mek (Inuit instrumental -nik, -mik), Siberian (and partly Alaskan)
Yupik -neng, -meng, Inuit -nin, -min (eastern -nit, -mit); prosecutive -kun (eastern
Inuit sg. -kkut); equative -tun (eastern Inuit -tut). Demonstrative and interrogative
adverbs have partly different allomorphs for the allative, locative, ablative and pros-
ecutive, most of them like the Aleut adverbial suffixes.

The allative sg. suffix -mun is found as Aleut -mud- in the type at-mud-
agan 'downwards' (2.1.6.3.3.), cf. Inupiaq at-mun 'downwards', parallel with the
type wa-ngud-agan 'in this direction, toward here' (2.1.7.9.1.), derived from the
adverbial locative (allative) A wa-ngus, E wa-ngun 'here, to here', CAY wa-vet,
CSY wha-vek, Inuit uvunga, uunga.

The adverbial forms qila-m 'in the morning, this morning', qila-ga-n 'to-
morrow' of the temporal noun qila- 'morning' (2.1.1.4.2.) are clearly relics of the
nominal locative which through the apocope of the final i (see 1986:98) merged

342 — Aleut Grammar — Aleut Compared With Eskimo

with the relative, cf. Inupiaq uvlaa-mi 'early in the morning', CSY unaa-mi 'to-
morrow, the next day', Greenl. aqagu-a-ni 'the following day', etc., and Aleut il-a-n,
Esk. ilu-a-ni 'at its inside, inside it'.

The -m of the multiplicative suffix -(i)di-in in ataqa-dim 'once', a(a)lg-idim,
also a(a)lg-im 'twice' , etc. (2. 1 .5.3.), may possibly reflect the nominal ablative / instru-
mental, cf. CSYataasi-meng, Greenl. ataatsi-mik 'once', etc. (discussion in 1989:34).

The adverbial ablative forms of temporal nouns, e.g. E amg-aan, A amg-
aax (-aagan) 'by night, at night' (2.1.1 .4.2.), are like the ablative of respectively the
3 A sg. of positional nouns (2.1.6.1.2.) and demonstrative adverbs (2.1.7.4.), which
probably reflect prosecutive forms, cf. Greenlandic unnua-kkut 'at night', etc.

With the loss of the oblique cases of ordinary nouns and pronouns, the Aleut
sentence structure was deeply transformed.

4.2. Object constructions

Eskimo has two well-known object constructions: (i) the so-called ergative
construction, viz. a verb with suffixal reference both to a definite object, possibly a
noun in the absolutive case (the same case as that of the nominal subject of an in-
transitive verb), and to an agent, possibly a noun in the relative case; (ii) the con-
struction sometimes called half-transitive or antipassive, viz. a verb with suffixal
reference only to a subject, possibly a noun in the absolutive case, and an indefinite
or only partly affected object in the ablative (Yupik) or instrumental (Inuit). E.g.
Central Alaskan Yupik (Reed etal. 1977,Miyaoka 1975, 1996)(i)angute-mtaqukaq
tangrr-a-a 'the man sees the bear' vs. (ii) angun taquka-mek tanger-tuq 'the man
sees a bear'; (i) (angutem) atsa-t ner-a-i 'he (the man) is eaUng the berries' vs. (ii)
(angun) atsa-nek ner'-uq 'he (the man) is eating berries'; (i) annga-ma angyaqa
atur-a-a 'my older brother is using my boat' vs. (ii) anngaqa angya-m-nek atur-
tuq 'my older brother is using a boat of mine'. So-called agentive verbs, like these
ones, are used in both constructions, while non-agentive verbs have in construction
(ii) a so-called half-transitive derivative (see 4.4.4.), e.g. (i) ama-m kuvyaq allg-a-a
'the woman tore the net' vs. (ii) amaq kuvya-mek allg-i-uq 'the woman tore a net' .

Both kinds of objects may have a reflexive suffix in reference to a third per-
son subject, to a subject (agent) in the relative case as well as to a subject in the
absolutive case, e.g. CAY (i) (angutem) kuv-a-a qalta-ni 'he (the man) spilled his
own pail (dehberately)' vs. (ii) (angun) kuv-i-uq qalta-mi-nek 'he (the man) spilled
his own pail (accidentally)' (Miyaoka 1996:344).

The nominal terms are obligatory in neither construction, cf. (i) tangrr-a-a
'(s)he sees it/him/her', (ii) tanger-tuq '(s)he sees something'; (i) ner-a-a '(s)he is
eating it', (ii)ner'-uq '(s)he is eating'; (i) allg-a-a '(s)he tore it', (ii) allg-i-uq '(s)he
tore something'; (i) kuv-a-a '(s)he spilled it', (ii) kuv-i-uq '(s)he spilled something'.

In Aleut there are two different possibilities: (a) a one-place verb with a speci-
fied object in the absolutive case, with a possible nominal subject also in the absolutive
case; (b) a two-place verb with anaphoric reference to a third person and a possible
nominal subject in the relative case, e.g. (a) taya|u-i qa-^ qa-ku-£ 'the man is

Aleut Compared With Eskimo — Aleut Grammar — 343

eating a/the fish', (b) taya|u-in qa-ku-u (< -a) 'the man is eating (or just ate) it'
(3.2.1-2.). In the case of (b), the referent of the anaphoric verb may be included in
the sentence as an outer subject, provided that it is separated from the verb by a
focused inner subject, normally a noun in the relative case (3.2.2.3.). Such sentences
are only superficially like an Eskimo structure with an initial object of a two-place
verb. In Eskimo the word order is relatively free and an object coming before the
subject (agent) remains an object, also with a reflexive suffix in reference to the then
following subject (agent), while in Aleut a reflexive suffix makes the object fully
specified and entails a one-place verb (3,2.5.1.1.), e.g. tayagu-^^ ada-an kidu-ku-i^
'the man is helping his own father' vs. CAY angute-m ata-ni ikayur-a-a.

Having a one-place verb, the Aleut construction (a) goes with the Eskimo
construction (ii), with the important difference that the Aleut object is in the absolutive
case rather than in an oblique case, lost in Aleut. Having a specified object, as op-
posed to the anaphoric construction (b), the Aleut construction (a) covers also the
Eskimo ergative construction (i) with a specified object in the absolutive case. This
leaves the anaphoric construction (b) as the Aleut reflex of the ergative construction,
e.g. tayagu-m kidu-ku-u 'the man is helping him/her', actually like CAY angute-m
ikayur-a-a. While in Eskimo the case marking of the functions of the nominal terms
admits a rather free word order, in Aleut the functional interrelation of the two terms
in the absolutive case is marked by the order SOV, e.g. Piitraii Paavila^ tugakui^
Teter hit Paul', Paavila^ Piitrai^ tugaku^ 'Paul hit Peter'. But a single nominal
term is case-marked: (a) Paavila-& tuga-ku-$ 'he hit Paul' vs. (b) Paavila-m tuga-
ku-u 'Paul hit him'.

Most importantly, while in Eskimo the specified as well as the anaphoric
object of the ergative construction is marked by the suffix of the verb, the reference
in the Aleut construction (a) is a zero-anaphora, as opposed to the suffixal reference
of the anaphoric construction (b).

4.3. First and second person arguments

The first and second persons are marked as subjects in much the same way in
Eskimo and in Aleut. The Eskimo endings of one-place verbs in the indicative and
participial moods are the cognates of the Aleut enclitic pronouns (2. 1 .2. 1 .2., 2. 1 .9, 1 . 1 „
etc.), with a more advanced stage of morphological fusion (cf. 1989: 15 f.), e.g. CAY
cali-u-ten, 'you are working', cali-u-nga 'I am working', Aleut awa-ku-:i-txin (A -t),
awa-ku-qing (A -q) < *-q-tking(a) (-i^ + ting). In dependent moods the subject
persons are marked by possessive suffixes in the relative case, e.g CAY cali-ku-vet,
Aleut A awa-gu-mis, E -min 'when/if you work'. In the former moods the endings
of two-place verbs with reference to a third person object (in Aleut anaphoric) are
possessive, e.g. CAY ikayur-a-qa 'I am helping him/her', ikayur-a-nka 'I am helping
them', like irnia-qa 'my child', irnia-nka 'my children' (irniaq 'child'), Aleut A
kidu-ku-ng, kidu-ku-ning, aniqdu-ng, aniqdu-ning. In dependent moods the sub-
ject is marked by possessive suffixes in the relative case followed by object suffixes,
CAY 3.p.sg. -ku/-gu, du. -kek, pi. -ki, old Atkan sg. -ka, pi. -kis, in Eastern infixed
(see 2.1.9.4.2.1., 2.1.9.8.).

344 — Aleut Grammar — Aleut Compared With Eskimo

In Eskimo, in accordance with the ergative system, the first and second per-
son object markers are in general the same as the subject endings of one-place verbs
in the indicative and participial, e.g. CAY angute-m ikayura-a-nga 'the man is
helping me', ikayur-a-m»ken 'I am helping you' (-ken is an allomorph of -ten, cf.
Aleut txin). In dependent moods they are on a par with the 3.p. object suffixes men-
tioned above, e.g. ikayu-qu-m-ken 'if I help you'. Also the "half-transitive" con-
strucUon is possible, with a free personal pronoun in the ablative (or instrumental)
case, e.g. CAY wangug-neng tangva-ku-neng aavurteciqu-t 'if they watch the
two of us, they will be amused' (Woodbury 1984:135).

In Aleut, the object markers for the first and second persons, as well as for
the reflexive third person, are free pronominal forms (2. 1.2.2.), which as fully speci-
fied in the relevant sense entail a one-place verb, e.g. taya§u-:S: ting kidu-ku-i^ 'the
man is helping me', txin kidu-ku-q(ing) 'I am helping you'. As objects of one-
place verbs they correspond to the Eskimo pronouns in the ablative (or instrumen-
tal) case, but through the Aleut transformation of the ergative system they may have
been dislocated from the suffixal position found in Eskimo (discussion in 1989:36
f.). Anyway, they are on a par with fully specified nouns (except for zero-anaphora,
see 3,9.2.1. and 3.11.2.6.1.) and differ fundamentally from the anaphoric third person
suffixes, as in tayagu-m kidu-ku-u 'the man is helping him', kidu-ku-ng 'I am
helping him', etc. Aleut has no non-reflexive third person pronouns like the Yupik
and Inupiaq ones (CAY ellii, CSY (e)llnga, I ilaa 'he, she, him, her', etc.), only demon-
strative pronouns, which have the syntactical function of nouns or determiners.

An Eskimo verb in the mood called appositional or subordinative (CAY) or
contemporative (Inuit), which like the Aleut conjunctive has mostly the same sub-
ject as the associated verb, has only one person suffix, in general either coreferential
with the subject of the associated verb or indicating an object. In accordance with
the ergative system, the first and second person suffixes are subjective or objective
depending on the context. The reflexive third person suffixes, which by themselves
indicate coreference with the subject of the associated (superordinate) verb, are natu-
rally subjective, but may also be coreferential with the reflexive object suffix of a
subordinate verb ( Greenlandic examples in Bergsland 1955:59, Fortescue 1984: 147).
The non-reflexive suffixes, sg. -ku, etc., are naturally objective, but in Alaskan Yupik
they may also, under certain condifions, be subjective in relation to an object (Miyaoka
1996:342; 1997:100ff). The subjective or objective function of these suffixes is in-
dependent of the transiUvity or intransitivity of the associated verb, e.g. CAY pissur-
yug-tu-a maligg-lu-ten 'I (-a < -nga) want to go hunting following you (-ten)',
nere-rrar-lu-ku an-ciq-uq 'first eating it he will go out' (Miyaoka 1997:72);
Greenlandic Kunuu-p ilaga-iu-git aullar-puq 'Kunuk (rel. subject of the
contemporative) being together with them (-git) went out' (Bergsland 1955:58; see
also Fortescue 1984:125).

The Aleut conjunctive has a subject marker, an enclitic pronoun, only when
used as a final predicate (2. 1 .9.2.) and with verbs of utterance and thought (3. 15.3. 1 .).
As a conjoined predicate (3.9.) with no complement or a fully specified one it has no

Aleut Compared With Eskimo — Aleut Grammar — 345

person suffix but in accordance with the Aleut system of refemnce has the suffixes -ka,
etc, in reference to an anaphoric complement. By zero-anaphora a specified nomi-
nal complement of the conjunctive may be shared by the following predicate (3.9.2. 1 .),
while an anaphoric reference of the conjunctive continues into the following verb, pos-
sibly by an enclitic dative (3.9.2.3.). The Aleut system of reference (specified vs.
anaphoric) thus applies to the whole sentence, possibly including both an outer and
an inner subject, while the Eskimo ergative system dominates single clauses.

4.4. Verbal valency

Eskimo verbs are either intransitive, admitting only one argument (a sub-
ject), or admit both one-place and two-place suffixally marked constructions, while
Aleut verbs are intransitive or transitive or both, a transitive verb having either a
specified or an anaphoric object (3.2.).

4.4. 1 . Intransitive verbs

Eskimo and Aleut verbs that admit only one suffixally marked argument,
possibly a nominal subject in the absolutive case, are, for example, CAY tai- and
Aleut haqa- *to come (to the area of the speaker)' ; tuqu-, asi^a- 'to die' (also 'death');
qavar-, saga- 'to sleep' (also nominal); qia-, qida- 'to cry' ; palu-, haaga- 'to starve' ;
qater-, quhma- 'to be white'; ange-, anguna- 'to be big, large'.

4.4.2. Agentive verbs

Eskimo agentive verbs may have the same subject in both constructions,
possibly a noun in the relative case in the ergative construction, the same noun in the
absolutive case in the "half-transitive" construction, while the corresponding Aleut
verbs are both transitive and intransitive, e.g. CAY nere-, Aleut qa- 'to eat'; nuteg-,
kalu- 'to shoot'. Here belong also, as a subtype, verbs with a possible local object,
e.g. CAY mayur-tuq 'he is going up', mayur-a-a 'he is climbing it', Aleut hanga-
(see 3.2.6.3.).

4.4.3. Eskimo non-agentive verbs

In accordance with the Eskimo ergative system these verbs may in both con-
structions have the same noun in the absolutive case, as an object in the ergative
construction and as a subject in the one-place construction. In the latter case, in
relation to the former, the verb is medio-passive or reflexive, e.g. CAY allg-a-a 'he
tears it', alleg-tuq 'it is torn (passive); it tears (medial)'; iir-a-a 'he hides it', iir-tuq
'it is hidden; it hides' ;erur-a-a 'he washes it', erur-tuq 'he washes himself (Miyaoka
1996:343).

In the case of the medio-passive subtype, the normal Aleut analogue of an
Eskimo non-agentive verb such as tamar- 'to lose' (tamartuq 'it is lost', tamaraa
'he lost it') is the pairhiki- 'to disappear; to get lost',hiki-t- 'to lose', the latter with
the transitivizing suffix -t- (stt Aleut Dictionary pp. 550 f ). The Eskimo non-agentive
verb kuve- 'to spill' (kuv'uq 'it spilled', kuvaa 'he spilled it') has an exceptional

Aleut Grammar
Unangam Tunuganaan Achixaasix

by

Knut Bergsland

© 1997 Alaska Native Language Center, University of Alaska Fairbanks;
and Knut Bergsland

Printed in the United States of America
All rights reserved

Library of Congress Cataloging-in-Publication data

Bergsland, Knut, 1914-

Aleut grammar = Unangam Tunuganaan Achixaasix / by Knut Bergsland.
p. cm. - (Alaska Native Language Center Research Papers ; no. 10)
Includes bibliographical references,
ISBN 1-55500-064-9 (alk. paper)

1. Aleut language— Granmiar. I. Title. H. Series: Alaska Native Language
Center Research Papers ; no. 10.
PM32,B46 1997

497'. 195~dc21 97-38814

CIP

Alaska Native Language Center

Research Paper Number 10

University of Alaska Fairbanks

1997

First Printing 1997 400 copies

Address correspondence to

Alaska Native Language Center
University of Alaska Fairbanks
RO. Box 757680
Fairbanks, AK 99775-7680